Key-Person Provisions: What Happens If the Manager Leaves

A key-person provision names individuals and attaches a consequence to their departure. In Nilgiri Growth Partners Fund II, invented, if both named people stop devoting substantially all of their business time to the fund, the investment period suspends automatically, with no vote, and no new investment may be made. Investors holding more than two thirds of commitments by value can approve a replacement and restart it. The provision has not been triggered.

Start with something entirely ordinary. In a small town the eldest son has been driving for a delivery firm for four years, and everybody in the house has watched the money arrive every month, so the household takes a loan to buy him a scooter. The loan is in the household's name. The bank's paperwork records an address, a co-signer and an income figure. Nowhere on that paperwork are the words the whole decision actually rested on: he is the one who drives. If he stops driving, every number in the file is still true and the reason for the loan has quietly disappeared.

A private fund has exactly this problem, at a much larger size and with a great deal more paper. Money is committed to a firm because of what a small number of people inside that firm have done before, and the contract that receives the money is signed by companies rather than by those people. A key-person provision is the single clause that closes that gap. The key-person clause is the one place in the whole set of documents where an individual human being is named, and the only place where a departure is written down as an event with a consequence attached.

What is a key-person provision actually protecting?

Look at what this fund's documents actually create. Nilgiri Growth Partners Fund II, invented, is settled as a trust. Nilgiri Trusteeship Services Private Limited, invented, is the trustee and holds the assets; Nilgiri Alternatives Advisors Private Limited, invented, is the investment manager and makes the investment decisions; Nilgiri Financial Holdings Private Limited, invented, is the sponsor and stands behind the manager. The trustee, the manager and the sponsor are three companies. In this fund the general partner's role is discharged by the manager and the trustee between them, and the contract is a trust deed and a contribution agreement rather than a partnership agreement, even though the vocabulary everybody uses at the table is the partnership one.

Now look at what the twelve investors were actually deciding when they committed Rs 4,90,00,00,000 between them. The twelve were not deciding that a private limited company with a registered office is a good thing to send money to for ten years. The decision was about people: who had run this kind of investment before, whose judgement had produced the outcome on the last vehicle, whose phone call gets returned when a company needs selling. The money was betting on judgement, and the contract can describe judgement least well of all.

So this fund names two people. Sundari Raghavan, an invented individual, is the managing partner of the manager and the first key personA named individual whose leaving triggers a consequence written into the fund's own contract.. Devendra Karnik, also invented, is a partner and the second. Their names sit in the fund's terms, and attached to those names is a condition about time: each must devote substantially all of their business time to the fund. The condition about time is the whole of what the clause holds onto. Not their titles, not their shareholding in the manager, not their reputation. Their working hours.

What happens when both of the named people go?

How Key-Person and Key-Manager Provisions Work

The clause of Nilgiri Growth Partners Fund II runs in the order the events would occur. Almost every mistake made about these clauses is a mistake about that order.

If both Sundari Raghavan and Devendra Karnik cease to devote substantially all of their business time to the fund during the investment periodThe stretch at the start of a fund's life during which it is allowed to buy new things. Afterwards it manages what it already has., the investment period suspends automatically. No vote is required. No notice has to be served by anybody on anybody in order to bring the stop about. From that moment no new investment may be made. The investment period resumes only if investors holding more than two thirds of commitments by value approve a replacement, and if they do not, the investment period ends permanently and the fund manages what it already holds until its term runs out.

The stop is the second thing that happens and the only decision is the fifth. That inversion is the whole design. The day two senior people walk out of a firm is not a day on which twelve institutional investors across three countries can be assembled to agree anything, so a clause that put the decision first would do nothing on the day it was most needed.

One more thing about the wording before the misreadings. The test is about time and attention, not about a job title and not about a resignation letter. Somebody who is still on the manager's website, still described as a partner, still attending the annual investor meeting, but who now spends most of the working week on a different vehicle, has ceased to devote substantially all of their business time to this fund. Whether that has happened is a question of fact, and this fund's own documents are what set the test.

Why does it work on its own instead of waiting for a vote?

Imagine a night watchman who has to raise the alarm before a fire alarm will sound. The alarm works perfectly except during a fire. A clause requiring investors to organise a vote in order to stop new investment has the same shape: it is useless at precisely the moment it exists for.

Think about what the day would actually look like. Two senior people leave a manager that has money still to deploy. Somebody has to notice. Somebody has to decide the departures count as a trigger. Somebody has to call twelve investors, several of whom are institutions where the person who reads the fund's letters is not the person who can cast a vote. Meanwhile the manager, whose fee keeps running and whose remaining people are still employed, is under no contractual bar at all and may keep buying companies with the money that is left. Every hour of that process is an hour in which the fund can commit capital that nobody would have approved if they had been asked first.

The automatic form removes all of it. Nobody has to be clever, nobody has to be fast, and nobody has to be brave enough to be the first investor to raise the question. The consequence attaches to the fact. The clause is written in the passive and in the present for that reason: the investment period suspends. The investment period is not suspended by anybody.

There is a second reason, and it is about who would otherwise carry the cost of acting. In a group of twelve investors of very different sizes, the small ones have every reason to wait for the large ones to move first, and the large ones have their own committees to satisfy. An automatic clause takes the coordination problem out of the contract entirely. Other funds carry the terms their own documents set.

Does one of the two leaving set it off?

No, and this is the second misreading. On this fund's terms the trigger needs both named people to cease devoting substantially all of their business time to the fund. One of the two going, however senior, however central to the last vehicle's outcome, does not suspend anything. The clause is written with the word both, and a clause written with the word either would be a different clause producing different results in three of the four possible states of the world.

A clause needing both names is worth sitting with rather than dismissing as a drafting weakness. A clause that trips on a single departure trips often: senior people move, retire, fall ill and take on other things, and a fund that suspends its buying every time one of them does is a fund that struggles to invest at all. A clause that needs both is a clause about the loss of the whole basis on which the money was committed, rather than about ordinary turnover. Whether that is the right line to draw is a drafting choice, and this invented fund drew it there.

What does a suspension stop, and what carries on regardless?

The third misreading costs the most. It produces a completely wrong picture of what the fund looks like the morning afterwards. A suspensionA stop on one activity while everything else the contract requires keeps running exactly as before. stops new investment. A suspension does not stop the fund.

Take the household example again. A house stops buying new furniture. The house does not stop paying the electricity bill, does not stop cooking, does not stop the children going to school, and does not sell the sofa it already has. Stopping one line of spending is not the same as shutting the house.

Nilgiri Growth Partners Fund II held five of its nine holdings at the end of its Year 9 Quarter 2, the record date every figure here is stated at. The five holdings need looking after whether or not anybody is allowed to buy a sixth. Each one needs board meetings attended, monthly information read, valuations struck, an administrator paid and an auditor answered. So the machinery keeps running, and the money to run it keeps being called.

The management fee is the item readers most often get wrong here, and it is worth being blunt about it. The fee does not stop because the investment period has paused. It is charged on the basis this fund's terms already set, and neither the basis nor the rate is altered by a key-person event. The manager's pay, and how the fee basis changes as the fund ages, is covered separately. Only the direction matters here: a suspension restricts the manager's freedom to buy and does not reduce the manager's pay.

The same is true of anything already signed. If the fund has committed to put money into a transaction under an agreement that already exists, the suspension does not tear that agreement up. And if a holding is sold during a suspension, the cash still goes out to investors in the order the fund's own terms set.

What does naming a role add that naming two people does not?

Key-Manager Provision

There is a second clause in the same area of the documents, and the two are related closely enough that readers routinely collapse them into one. A key-person clause names individuals. A key-manager clause attaches the same kind of consequence to a role inside the fund rather than only to a name.

Why would anybody bother writing both? Because of a gap that a name-only clause leaves open. Picture a small restaurant that is famous for one cook. The lease says the cook is named in it. The cook stays on the payroll, keeps the title, and stops cooking; a nephew now runs the kitchen. Every word of the lease is satisfied and the thing the customers came for has gone. A clause about the role asks a different question: is the head cook's job actually being done, and by somebody doing it properly?

Applied to a fund, the role version asks whether the senior investment role is genuinely filled and carries the same time and attention, rather than whether two particular names still appear somewhere. The test in both clauses turns on time for that reason: substantially all of a person's business time going to this fund is a fact about the working week, and a title is not.

The role clause is the wider of the two and the name clause is the sharper, and a fund that writes both is protected against a departure and against a hollowing out. Nilgiri Growth Partners Fund II carries both, and they stand as two clauses rather than one clause with two names. Terms differ from fund to fund.

How does the investment period start again, and who decides?

A suspension is not an ending. A suspension is a pause with one exit, and the exit needs somebody to walk through it. The manager may propose a replacement for the key persons who have gone, and investors holding more than two thirds of commitmentsThe amount each investor has contractually promised to the fund, whether or not it has yet been asked for. by value may approve that replacement. If they do, the investment period resumes and the fund may buy again. If they do not, the investment period ends permanently, and the fund spends the rest of its term managing and selling what it already holds.

Notice which of the two branches is the one that needs work. Resuming needs an affirmative decision by investors holding a stated share of the money, and ending permanently needs nothing at all. Investors who dislike the proposed replacement do not have to organise against it; investors who are simply slow, or distracted, or unable to get an internal approval through in time, land in the same place as investors who actively object. Inaction is not neutral here. Inaction is one of the two outcomes.

What does two thirds by value look like across twelve unequal investors?

Two thirds sounds like a description of a crowd. It is not. Two thirds by value describes money, and in this fund the money is distributed very unevenly across the twelve investors, so a sentence about two thirds of the investors and a sentence about two thirds of the commitments describe completely different groups of people.

Before any arithmetic, name the denominator. This fund has two of them and they are not interchangeable. The twelve investors committed Rs 4,90,00,00,000 between them. The manager committed a further Rs 10,00,00,000 of its own, so total commitments are Rs 5,00,00,00,000. A share of Rs 4,90,00,00,000 is a different number from a share of Rs 5,00,00,00,000, and any sentence about a threshold that does not say which base it is using has not finished.

Here are the twelve, largest first, with the running total kept in both bases.

Now the threshold. Two thirds of the Rs 4,90,00,00,000 the twelve investors committed is Rs 3,26,66,66,667 taken to the nearest rupee, and the clause needs more than that. The four largest hold Rs 1,00,00,00,000 plus Rs 80,00,00,000 plus Rs 75,00,00,000 plus Rs 60,00,00,000, adding to Rs 3,15,00,00,000. The four together are 64.3 per cent of the Rs 4,90,00,00,000 base and Rs 11,66,66,667 short of the threshold. The four largest investors in this fund hold nearly two thirds of it between them and still cannot restart the investment period on their own. Add the fifth largest at Rs 50,00,00,000 and the running total reaches Rs 3,65,00,00,000, being 74.5 per cent of that base, which clears it with room to spare.

Run it on the other base and the answer does not move. Two thirds of the Rs 5,00,00,00,000 of total commitments is Rs 3,33,33,33,333 to the nearest rupee; the four largest at Rs 3,15,00,00,000 are 63.0 per cent and Rs 18,33,33,333 short, and the five largest at Rs 3,65,00,00,000 are 73.0 per cent and clear. Five investors on either base. Five on either base is a useful property of this particular set of numbers rather than a general rule, and it is the reason the result holds without first fixing which base this fund's own documents use.

Now put the count of heads beside the count of rupees. The gap between the two counts is the entire point. Five investors out of twelve is 41.7 per cent of the heads and 74.5 per cent of the Rs 4,90,00,00,000. Go the other way and it is starker. Take every one of the twelve except investor 1 at Rs 1,00,00,00,000 and investor 3 at Rs 80,00,00,000. The ten smallest hold Rs 3,10,00,00,000 between them, and that is 63.3 per cent of the Rs 4,90,00,00,000 base and 62.0 per cent of the Rs 5,00,00,00,000 base. Ten of the twelve investors, being 83.3 per cent of the heads, still do not clear two thirds by value. A vote counted by head and a vote counted by value are not approximations of each other in this fund. The two counts can point in opposite directions.

The tail is worth stating on its own for the same reason. Investors 7 to 12 hold Rs 25,00,00,000, Rs 20,00,00,000, Rs 15,00,00,000, Rs 10,00,00,000, Rs 10,00,00,000 and Rs 5,00,00,000, adding to Rs 85,00,00,000. The tail is 17.0 per cent of the Rs 5,00,00,00,000 of total commitments and 17.3 per cent of the Rs 4,90,00,00,000 of investor commitments. Six investors, being exactly half the heads, carrying about a sixth of the money. Meanwhile investor 1 alone holds Rs 1,00,00,00,000, being 20.0 per cent of total commitments and 20.4 per cent of investor commitments, so one investor is worth more than the smallest six put together.

How is any of this different from removing the manager?

Completely different, and this is the sharpest distinction of the whole subject. A key-person provision and a removal provisionA clause letting investors vote the manager out of the job, on grounds the contract states. are two separate clauses in this fund's documents, they answer two different questions, and they work in opposite ways. Letting one stand in for the other is the fastest way to be wrong about what a set of fund terms actually does.

The key-person clause is automatic and it is about departure. Nobody decides. The key-person clause asks one question: are the two named people still there doing the work? The consequence is a stop on buying.

The removal clause is a vote and it is about conduct or performance. Somebody has to call it and somebody has to count it, and the question is a completely different one: should this manager still be the manager? Removal comes in two flavours. For causeOn specified grounds, such as a breach of the agreement or misconduct. These have to be shown before the vote can be used. needs stated grounds such as a breach or misconduct established first. Without cause needs no grounds at all, and the bar is set instead by the majority required. The fund's own documents set the grounds and the majority for each.

Four years of poor performance with nobody leaving does not touch the key-person clause, and two departures with nothing wrong at all does not touch the removal clause. They are not two settings on one dial. Nilgiri Growth Partners Fund II, invented, has both of them and neither has been used.

Which of the two committees can actually approve an investment?

LPAC vs Investment Committee

Two bodies sit near this fund's decisions and readers mix them up more often than they mix up anything else in the subject. One is the investment committeeThe manager's own decision-making body. It signs off each purchase and each sale the fund makes.. The other is the investor advisory committeeA committee drawn from the fund's investors that consents to a defined list of matters and decides nothing else., called the limited partner advisory committee (LPAC) in the shorthand many documents use.

The investment committee has five members. Four of them come from the manager, Nilgiri Alternatives Advisors Private Limited, invented, and one is external. The investment committee approves every investment the fund makes and every realisationA sale, a partial sale or any other event that turns a holding back into cash. the fund takes. The committee is the manager's own body and it answers to the manager.

The investor advisory committee has seven members drawn from the investors themselves, and the investor seats are exactly what makes the committee easy to mistake for a board. The committee consents on four matters: conflicts, valuation policy, the first extension of the fund's term, and any change to the investment policy. Who sits on it, how the seats were allocated and what each of those four consent matters involves in detail are covered separately. The absence matters more than the four consent matters do. Approving an investment is not one of the four things the investor advisory committee consents to, so it can neither approve a purchase nor reject one.

Think of a housing society. The managing committee decides which contractor repaints the building. A residents' consultation group is asked before the society changes its own accounting rules or grants an exception to one flat. If the paint job goes badly, the consultation group did not choose the contractor and could not have refused him. Being drawn from the residents does not make a body the one that spends the money.

How does somebody actually use this clause in a working week?

Take an analyst inside a fund investor, one of the twelve, whose job is to read the terms of vehicles the institution has already committed to and report anything that has moved. The key-person clause reaches that analyst in three ordinary ways, none of which involves a crisis.

The first is a reading job, done once. The analyst finds the clause, writes down the two names, writes down the test attached to those names, writes down the trigger word, both rather than either, and writes down the threshold and the base it is measured against. Five facts, and the fifth one is where most notes are wrong. A note recording two thirds without recording two thirds of what cannot be used later.

The second is a monitoring job, done continuously and mostly by noticing. Managers announce new vehicles, senior people take on outside roles, and firms reorganise. None of that is a trigger by itself. The question the analyst keeps asking is narrower: are these two people still putting substantially all of their business time into this fund? The clause is written about working hours, so the thing worth watching is where the working hours are going, not who has resigned.

The third arrives only if the clause ever triggers, and it is a calendar job rather than an analytical one. Because the resuming branch is the one that needs a positive decision, an investor who wants the fund to keep buying has to get an internal approval through in time to be counted. Doing nothing lands in the permanent-ending branch, so even an investor who is indifferent still has to decide. The asymmetry is why the analyst's note is worth writing before anything happens rather than after.

The same reading serves somebody on the other side of the table, inside a manager, for a different reason: the clause tells them exactly which two working diaries the fund's documents are watching, and that everything else about seniority and titles is outside the test.

Has this fund's provision ever been triggered?

No. Nilgiri Growth Partners Fund II, invented, reached the end of its Year 9 Quarter 2 with its key-person provision never having been triggered. Sundari Raghavan and Devendra Karnik both remain, both are still devoting substantially all of their business time to the fund, the investment period ran to its own contracted end rather than being suspended, and no vote on any replacement has ever been called.

Every sequence, branch and threshold set out above is therefore a counterfactual: it is what this fund's own clause would do, not anything it has done. A clause that has sat unused in a document for nine years is in the ordinary condition of most clauses of this kind, and the fact that it has never fired says nothing at all about whether the clause was worth having.